Method and apparatus for determining channel state information

1. A method for determining channel state information in an orthogonal frequency division multiplexing (OFDM) transmission scheme in a receiver, comprising:

estimating a frequency selective interference noise covariance matrix per resource block, wherein a resource block comprises an aggregate of resource elements that are modulation structures that comprise one subcarrier by one symbol;

obtaining a normalized channel matrix by weighting a channel matrix using the estimated frequency selective interference noise covariance matrix as an interference estimate;

determining a channel equivalent matrix by multiplying the normalized channel matrix with all available precoding matrices;

computing a post-equalization signal-to-interference plus noise ratio (SINR) as a function of an equalizer in use in the receiver using the determined channel equivalent matrix;

deriving mutual information for each post-equalization SINR as a function of the modulation used;

averaging the mutual information in sub-bands; and

deriving channel state information from the averaged mutual information.

2. The method of claim 1 , wherein deriving the mutual information comprises:

computing a hypothetical capacity for all computed post-equalization SINR; and

applying a non-linear transform to the hypothetical capacity, which non-linear transform models an actual modulation type and a receiver specific QAM demapper implementation, to obtain the mutual information for each post-equalization SINR value.

3. The method of claim 1 , wherein deriving channel state information from the averaged mutual information comprises:

selecting a precoding matrix index (PMI) for each possible number of code words such that it maximizes a sum of spectral efficiency over all code words.

4. The method of claim 3 , wherein deriving channel state information from the averaged mutual information further comprises:

selecting a rank indication (RI) of a channel such that the sum of spectral efficiency over all available code words is maximized, with the a priori knowledge of the PMI.

5. The method of claim 3 , wherein deriving channel state information from the averaged mutual information further comprises:

selecting a channel quality indication (COI) as a maximum achievable spectral efficiency on each code word.

6. The method according to claim 1 , wherein the equalizer uses maximum ratio combining (MRC), and computing a post-equalization SINR comprises determining a squared Frobenius norm of the channel matrix.

7. The method according to claim 1 , wherein the equalizer uses MIMO minimum mean square error (MMSE) equalization, and computing a post-equalization SINR comprises computing the channel covariance matrix and deriving therefrom the SINR for each code word.

8. The method according to claim 1 , wherein the equalizer uses MIMO maximum likelihood (ML) detection, and computing a post-equalization SINR comprises deriving the post-equalization SINR from eigenvalues of the modified channel covariance matrix.

9. An orthogonal frequency division multiplexing (OFDM) baseband receiver, comprising:

a noise level estimation unit configured to estimate a noise level and channel conditions from received symbols;

a demodulator unit configured to demodulate and decode data comprised of the received symbols,

an estimation unit configured to estimate time and frequency offset parameters based on the received symbols; and

a feedback information estimation unit configured to estimate feedback information using the noise level and channel condition estimates, wherein the feedback information estimation unit comprises:

a de-precoding unit configured to determine a channel equivalent matrix by multiplying a normalized channel matrix with all available precoding matrices;

a post-equalization SINR unit configured to compute a post-equalization signal-to-interference plus noise ratio (SINR) as a function of an equalizer in use in the receiver;

a mutual information derivation unit configured to derive mutual information for each post-equalization SINR as a function of the modulation used;

an averaging unit configured to average the mutual information in sub-bands; and

a channel state information derivation unit configured to derive channel state information from the averaged mutual information.

10. The baseband receiver of claim 9 , wherein the mutual information derivation unit comprises a computing unit configured to compute a hypothetical capacity for all code words, and wherein the mutual information derivation unit comprises a mutual information calculation block configured, for each post-equalization SINR, to apply a non-linear transform to the hypothetical capacity, which non-linear transform models an actual modulation rate and the receiver specific QAM demapper implementation.

11. The baseband receiver of claim 9 , wherein the channel state information derivation unit is operable to select a precoding matrix index (PMI) for each possible number of code words such that it maximizes a sum of spectral efficiency over all code words.

12. The baseband receiver of claim 11 , wherein the channel state information derivation unit is further operable to select a rank indication (RI) of the channel such that the sum of spectral efficiency over all available code words is maximized, with the a priori knowledge of the PMI.

13. The baseband receiver of claim 11 , wherein the channel state information derivation unit is further operable to select a channel quality indication (CQI) as the maximum achievable spectral efficiency on each code word.